Prevalence and determinants of rheumatic heart disease among school-going children in Dhanusha district, southern Nepal: a cross-sectional echocardiographic screening study

Rheumatic heart disease (RHD) is a leading preventable cause of cardiac death in children in low and middle-income countries. Nepal’s epidemiological data come mainly from auscultation surveys that miss subclinical disease, and no echocardiographic screening study had been conducted in Dhanusha district, a densely populated, low-income region in southern Nepal. We aimed to determine the prevalence of borderline and definite RHD among school children (6–16 years) in Dhanusha using the 2012 World Heart Federation (WHF) echocardiographic criteria, identify independent predictors, and quantify school-level clustering via the intraclass correlation coefficient (ICC). In a cross-sectional study (January 2023–December 2024), we screened 4,536 children from 8 public schools selected by four-stage cluster sampling. RHD was classified by WHF 2012 criteria; predictors were identified using random-effects logistic regression with school as random intercept. Ethical approval was from the Nepal Health Research Council (Protocol No. 155/2023). Overall prevalence of borderline or definite RHD was 18.7 per 1,000 (95% CI 15.1–23.0); definite RHD was 6.8 per 1,000 (95% CI 4.7–9.7) and borderline RHD 11.9 per 1,000 (95% CI 9.0–15.5). Prevalence was higher in girls (23.3 per 1,000) than boys (13.6 per 1,000; P=0.02), with the peak in girls aged 10–14 years (26.0 per 1,000). Subclinical disease accounted for 64.7% of cases; auscultation sensitivity was 35.3%. Mitral valve involvement predominated. Female sex was the sole independent predictor (OR 1.60, 95% CI 1.02–2.53; P=0.043). The school-level ICC was 0.19 (95% CI 0.07–0.44; P<0.001), giving a design effect of ≈109. The echocardiographic RHD burden in Dhanusha (18.7 per 1,000) is the highest documented in Nepal. Two-thirds of cases are subclinical. Female sex and school attended explain a similar amount of variance in RHD risk, supporting school-targeted screening and informing sample size planning for future cluster-based surveillance.